A repo of examples of how Sage can be used with Weaviate and other VectorDBs for cool machine-learning related tasks.
| Title | Language | Description |
|---|---|---|
| Text/Image search using CLIP | Python | Use text or images to search through Sage images using CLIP (multi2vec-clip Weaviate module). |
| MultiModal search using ImageBind | Python | Use text,images,audio, & more to search through Sage images using ImageBind (multi2vec-bind Weaviate module). |
| MultiModal Hybrid search using ImageBind and Gemma 3 | Python | Use text and images to search through Sage images using DFN Clip. Images are also captioned using Gemma 3 to conduct a Hybrid search where a keyword search and a vector search is conducted. WEAVIATE v4 IS USED HERE |
| Sage Chat | Python | This implements an AI agent that can use multiple functions including MultiModal Hybrid search using ImageBind and Florence 2. WEAVIATE v4 IS USED HERE |
| INQUIRE Benchmark | Python | This scores our MultiModal Hybrid search using INQUIRE. WEAVIATE v4 IS USED HERE |
| TODO Vectordb on the edge | Python | This example runs a vector database on the edge using nvidia's nanodb. |
| TODO Monitoring Setup with Prometheus & Grafana | yaml | This example does not describe any use case, but rather shows a way of how to start, operate and configure Weaviate with Prometheus-Monitoring and a Grafana Instance with some sample dashboards. |